----------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_logs/appendix_table2.log
  log type:  text
 opened on:  24 Jun 2022, 06:24:12

. 
.         use "$Rep_smokelabor/1_build/regdata/county_quarter.dta", clear 

.         
.         local tbl_settings_log format(%6.3f) parentheses(stderr) asterisk()

.         local tbl_settings_pmil format(%6.1f) parentheses(stderr) asterisk()

.         
.         ** panel a 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_dens) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros
> )
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    18.26
                                                      Prob > F      =   0.0000
Total (centered) SS     =   8762876942                Centered R2   =  -0.0757
Total (uncentered) SS   =   8762876942                Uncentered R2 =  -0.0757
Residual SS             =   9426576610                Root MSE      =    356.7

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -83.40949    19.5169    -4.27   0.000    -121.6894   -45.12955
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             53.153
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3571.413
                         (Kleibergen-Paap rk Wald F statistic):         89.854
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_dens
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_a.dta", addlabel(KleibergenPaap_F, `e(rk
> f)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_a.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_a.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_dens) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyfip f
> e_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    19.47
                                                      Prob > F      =   0.0000
Total (centered) SS     =  1.77722e+13                Centered R2   =  -0.0151
Total (uncentered) SS   =  1.77722e+13                Uncentered R2 =  -0.0151
Residual SS             =  1.80409e+13                Root MSE      =    15604

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1738.897   394.1136    -4.41   0.000    -2511.901   -965.8933
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             53.010
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3581.904
                         (Kleibergen-Paap rk Wald F statistic):         89.340
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_dens
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_a.dta", table(employment, `tbl_settings_
> pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_a.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_dens) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =     6.09
                                                      Prob > F      =   0.0137
Total (centered) SS     =  6.03036e+12                Centered R2   =  -0.0034
Total (uncentered) SS   =  6.03036e+12                Uncentered R2 =  -0.0034
Residual SS             =  6.05078e+12                Root MSE      =     9009

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -506.2616   205.1053    -2.47   0.014    -908.5497   -103.9736
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             53.130
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3544.936
                         (Kleibergen-Paap rk Wald F statistic):         89.600
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_dens
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           0         637     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_a.dta", table(lfp, `tbl_settings_pmil') 
> append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_a.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_a.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +-------------------------------------------------------+
     |              var     earnings   employment        lfp |
     |-------------------------------------------------------|
  1. | KleibergenPaap_F       89.854                         |
  2. |                                                       |
  3. |        pm25_coef   -83.409***   -1738.9***   -506.3** |
  4. |      pm25_stderr     (19.517)      (394.1)    (205.1) |
  5. |                                                       |
  6. |     observations       74,725       74,725     75,193 |
     +-------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_a.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_a.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel b, first row
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll g*_tavg g*_wdir (pm25=hms_deep) [aw=seer_pop], a(g_ppt g_wspd fe_countyqtroy fe_styr)
>  cluster(countyfip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 16 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74699
Number of clusters (fe_stqtros) =   2520              F( 16,  1684) =     5.26
                                                      Prob > F      =   0.0000
Total (centered) SS     =   8719824620                Centered R2   =  -0.1734
Total (uncentered) SS   =   8719824620                Uncentered R2 =  -0.1734
Residual SS             =  1.02322e+10                Root MSE      =    371.8

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -124.0738   28.96657    -4.28   0.000     -180.888    -67.2595
     g1_tavg |  -4.678285   2.169299    -2.16   0.031    -8.933091   -.4234792
     g2_tavg |   4.250758   3.908407     1.09   0.277    -3.415089     11.9166
     g3_tavg |  -3.394949   1.607533    -2.11   0.035    -6.547921   -.2419762
     g4_tavg |  -3.935323    1.25586    -3.13   0.002    -6.398534   -1.472111
     g5_tavg |   -3.04691   1.323454    -2.30   0.021    -5.642698   -.4511226
     g6_tavg |  -3.837706     1.2586    -3.05   0.002    -6.306292   -1.369121
     g7_tavg |    .038658    .932826     0.04   0.967    -1.790962    1.868278
     g8_tavg |   2.201118   .8388081     2.62   0.009     .5559015    3.846334
     g9_tavg |   5.097291   1.237445     4.12   0.000       2.6702    7.524382
     g1_wdir |   1.208916   6.892954     0.18   0.861    -12.31074    14.72857
     g2_wdir |   3.028031   6.955773     0.44   0.663    -10.61484     16.6709
     g3_wdir |   .1387737   6.984636     0.02   0.984    -13.56071    13.83826
     g4_wdir |    3.32187   7.474377     0.44   0.657    -11.33818    17.98192
     g5_wdir |   4.931778   7.528241     0.66   0.512    -9.833916    19.69747
     g6_wdir |   .6385314   7.393369     0.09   0.931    -13.86263    15.13969
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             36.978
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             1861.809
                         (Kleibergen-Paap rk Wald F statistic):         45.885
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Included instruments: g1_tavg g2_tavg g3_tavg g4_tavg g5_tavg g6_tavg g7_tavg
                      g8_tavg g9_tavg g1_wdir g2_wdir g3_wdir g4_wdir g5_wdir
                      g6_wdir
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
          g_ppt |        10           0          10     |
         g_wspd |        10           1           9     |
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           1         630    ?|
--------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave pm25 using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta", addlabel(KleibergenPaap_F,
>  `e(rkf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal g*_tavg g*_wdir (pm25=hms_deep) [aw=seer_pop16plus], a(g_ppt g_wspd fe_countyqtroy
>  fe_styr) cluster(countyfip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 16 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74699
Number of clusters (fe_stqtros) =   2520              F( 16,  1684) =    13.18
                                                      Prob > F      =   0.0000
Total (centered) SS     =  1.77381e+13                Centered R2   =  -0.0122
Total (uncentered) SS   =  1.77381e+13                Uncentered R2 =  -0.0122
Residual SS             =  1.79536e+13                Root MSE      =    15573

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   -2065.65    544.191    -3.80   0.000    -3133.012   -998.2883
     g1_tavg |  -224.0052   91.90697    -2.44   0.015    -404.2692   -43.74132
     g2_tavg |  -144.2463   91.77041    -1.57   0.116    -324.2424    35.74975
     g3_tavg |  -257.7393   66.59499    -3.87   0.000    -388.3569   -127.1216
     g4_tavg |  -266.3157   67.17907    -3.96   0.000     -398.079   -134.5525
     g5_tavg |  -167.6051   59.35632    -2.82   0.005    -284.0251   -51.18522
     g6_tavg |  -133.1392   57.64317    -2.31   0.021     -246.199   -20.07937
     g7_tavg |  -48.73597   54.96107    -0.89   0.375    -156.5352    59.06324
     g8_tavg |   43.45379   57.95475     0.75   0.453    -70.21713    157.1247
     g9_tavg |     157.91   62.44576     2.53   0.012     35.43051    280.3895
     g1_wdir |  -6.373358     179.04    -0.04   0.972    -357.5378    344.7911
     g2_wdir |    36.6388    184.469     0.20   0.843    -325.1739    398.4515
     g3_wdir |  -149.9594   186.1018    -0.81   0.420    -514.9747    215.0559
     g4_wdir |  -137.7691   188.2434    -0.73   0.464    -506.9848    231.4466
     g5_wdir |  -21.12116   187.7224    -0.11   0.910    -389.3149    347.0725
     g6_wdir |  -167.2653    192.739    -0.87   0.386    -545.2985    210.7679
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             36.783
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             1863.628
                         (Kleibergen-Paap rk Wald F statistic):         45.365
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Included instruments: g1_tavg g2_tavg g3_tavg g4_tavg g5_tavg g6_tavg g7_tavg
                      g8_tavg g9_tavg g1_wdir g2_wdir g3_wdir g4_wdir g5_wdir
                      g6_wdir
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
          g_ppt |        10           0          10     |
         g_wspd |        10           1           9     |
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           1         630    ?|
--------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave pm25 using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta", table(employment, `tbl_set
> tings_pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp g*_tavg g*_wdir (pm25=hms_deep) [aw=seer_pop] , a(g_ppt g_wspd fe_countyqtroy fe_styr) 
> cluster(countyfip fe_stqtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 16 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75167
Number of clusters (fe_stqtros) =   2548              F( 16,  1685) =    11.22
                                                      Prob > F      =   0.0000
Total (centered) SS     =  6.01746e+12                Centered R2   =  -0.0020
Total (uncentered) SS   =  6.01746e+12                Uncentered R2 =  -0.0020
Residual SS             =  6.02930e+12                Root MSE      =     8996

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -814.8175   231.4948    -3.52   0.000    -1268.865     -360.77
     g1_tavg |  -58.16321   46.89595    -1.24   0.215    -150.1437    33.81723
     g2_tavg |   -14.0774   53.79898    -0.26   0.794    -119.5973    91.44247
     g3_tavg |  -91.53008    34.4245    -2.66   0.008    -159.0494    -24.0108
     g4_tavg |  -29.57221   31.55113    -0.94   0.349    -91.45573    32.31131
     g5_tavg |   -60.0679   27.20221    -2.21   0.027    -113.4216   -6.714217
     g6_tavg |  -69.71749   21.78721    -3.20   0.001    -112.4503   -26.98465
     g7_tavg |  -40.03145   23.00759    -1.74   0.082    -85.15792    5.095009
     g8_tavg |   3.802453   22.98447     0.17   0.869    -41.27866    48.88356
     g9_tavg |  -18.68133   24.86973    -0.75   0.453    -67.46014    30.09748
     g1_wdir |  -768.9019   86.90241    -8.85   0.000    -939.3499   -598.4539
     g2_wdir |  -775.0545   87.69855    -8.84   0.000    -947.0641    -603.045
     g3_wdir |  -788.3105   82.94917    -9.50   0.000    -951.0047   -625.6162
     g4_wdir |  -806.8457   87.66498    -9.20   0.000    -978.7894   -634.9019
     g5_wdir |  -779.6135   96.94353    -8.04   0.000    -969.7559   -589.4711
     g6_wdir |  -806.6353     87.793    -9.19   0.000    -978.8301   -634.4405
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             36.480
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             1823.177
                         (Kleibergen-Paap rk Wald F statistic):         45.188
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Warning: estimated covariance matrix of moment conditions not of full rank.
         overidentification statistic not reported, and standard errors and
         model tests should be interpreted with caution.
Possible causes:
         number of clusters insufficient to calculate robust covariance matrix
         singleton dummy variable (dummy with one 1 and N-1 0s or vice versa)
partial option may address problem.
------------------------------------------------------------------------------
Instrumented:         pm25
Included instruments: g1_tavg g2_tavg g3_tavg g4_tavg g5_tavg g6_tavg g7_tavg
                      g8_tavg g9_tavg g1_wdir g2_wdir g3_wdir g4_wdir g5_wdir
                      g6_wdir
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
          g_ppt |        10           0          10     |
         g_wspd |        10           1           9     |
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           1         636    ?|
--------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave pm25 using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta", table(lfp, `tbl_settings_p
> mil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +---------------------------------------------------------+
     |              var      earnings   employment         lfp |
     |---------------------------------------------------------|
  1. | KleibergenPaap_F        45.885                          |
  2. |                                                         |
  3. |        pm25_coef   -124.074***   -2065.7***   -814.8*** |
  4. |      pm25_stderr      (28.967)      (544.2)     (231.5) |
  5. |                                                         |
  6. |     observations        74,699       74,699      75,167 |
     +---------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b1.dta saved
.                 restore 
.                         
.         }

.         
.         ** panel b, second row
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(c.g1_wdir##i.statefip c.g2_wdir##i.statefip c.g3_wdi
> r##i.statefip c.g4_wdir##i.statefip c.g5_wdir##i.statefip c.g6_wdir##i.statefip fe_countyqtroy fe_styr) cluster(countyfip fe_stq
> tros)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 177 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    19.42
                                                      Prob > F      =   0.0000
Total (centered) SS     =   8561813089                Centered R2   =  -0.1531
Total (uncentered) SS   =   8561813089                Uncentered R2 =  -0.1531
Residual SS             =   9872292877                Root MSE      =    365.7

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -113.7938   25.82424    -4.41   0.000    -164.4448   -63.14281
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             43.522
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2229.627
                         (Kleibergen-Paap rk Wald F statistic):         56.035
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
------------------------------------------------------------+
        Absorbed FE | Categories  - Redundant  = Num. Coefs |
--------------------+---------------------------------------|
           statefip |        49           0          49     |
 statefip#c.g1_wdir |        49           0          49    ?|
           statefip |        49          49           0     |
 statefip#c.g2_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g3_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g4_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g5_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g6_wdir |        49           0          49    ?|
     fe_countyqtroy |      6676        6676           0    *|
            fe_styr |       631          49         582    ?|
------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(c.g1_wdir##i.statefip c.g2_wdir##i.statefip
>  c.g3_wdir##i.statefip c.g4_wdir##i.statefip c.g5_wdir##i.statefip c.g6_wdir##i.statefip fe_countyqtroy fe_styr) cluster(countyf
> ip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 178 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    13.16
                                                      Prob > F      =   0.0003
Total (centered) SS     =  1.73209e+13                Centered R2   =  -0.0141
Total (uncentered) SS   =  1.73209e+13                Uncentered R2 =  -0.0141
Residual SS             =  1.75650e+13                Root MSE      =    15428

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1643.878   453.1511    -3.63   0.000    -2532.677   -755.0792
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             43.415
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2233.492
                         (Kleibergen-Paap rk Wald F statistic):         55.469
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
------------------------------------------------------------+
        Absorbed FE | Categories  - Redundant  = Num. Coefs |
--------------------+---------------------------------------|
           statefip |        49           0          49     |
 statefip#c.g1_wdir |        49           0          49    ?|
           statefip |        49          49           0     |
 statefip#c.g2_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g3_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g4_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g5_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g6_wdir |        49           0          49    ?|
     fe_countyqtroy |      6676        6676           0    *|
            fe_styr |       631          49         582    ?|
------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(c.g1_wdir##i.statefip c.g2_wdir##i.statefip c.g3_wdir
> ##i.statefip c.g4_wdir##i.statefip c.g5_wdir##i.statefip c.g6_wdir##i.statefip fe_countyqtroy fe_styr) cluster(countyfip fe_stqt
> ros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 177 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =    14.41
                                                      Prob > F      =   0.0002
Total (centered) SS     =  5.88794e+12                Centered R2   =  -0.0096
Total (uncentered) SS   =  5.88794e+12                Uncentered R2 =  -0.0096
Residual SS             =  5.94474e+12                Root MSE      =     8947

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -758.3189   199.7392    -3.80   0.000    -1150.082   -366.5559
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             43.271
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2194.708
                         (Kleibergen-Paap rk Wald F statistic):         55.675
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
------------------------------------------------------------+
        Absorbed FE | Categories  - Redundant  = Num. Coefs |
--------------------+---------------------------------------|
           statefip |        49           0          49     |
 statefip#c.g1_wdir |        49           0          49    ?|
           statefip |        49          49           0     |
 statefip#c.g2_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g3_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g4_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g5_wdir |        49           0          49    ?|
           statefip |        49          49           0    ?|
 statefip#c.g6_wdir |        49           0          49    ?|
     fe_countyqtroy |      6677        6677           0    *|
            fe_styr |       637          49         588    ?|
------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +---------------------------------------------------------+
     |              var      earnings   employment         lfp |
     |---------------------------------------------------------|
  1. | KleibergenPaap_F        56.035                          |
  2. |                                                         |
  3. |        pm25_coef   -113.794***   -1643.9***   -758.3*** |
  4. |      pm25_stderr      (25.824)      (453.2)     (199.7) |
  5. |                                                         |
  6. |     observations        74,725       74,725      75,193 |
     +---------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b2.dta saved
.                 restore 
.                         
.         }

.         
.         ** panel b, third row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(c.g1_wdir##i.countyfip c.g2_wdir##i.countyfip c.g3_w
> dir##i.countyfip c.g4_wdir##i.countyfip c.g5_wdir##i.countyfip c.g6_wdir##i.countyfip fe_countyqtroy fe_styr) cluster(countyfip 
> fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 554 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    17.81
                                                      Prob > F      =   0.0000
Total (centered) SS     =   7625345457                Centered R2   =  -0.1758
Total (uncentered) SS   =   7625345457                Uncentered R2 =  -0.1758
Residual SS             =   8965936053                Root MSE      =    374.3

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -123.4159   29.24696    -4.22   0.000    -180.7801   -66.05163
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             44.229
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2019.453
                         (Kleibergen-Paap rk Wald F statistic):         50.023
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------------+
         Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------------+---------------------------------------|
           countyfip |      1685        1685           0    *|
 countyfip#c.g1_wdir |      1685           2        1683    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g2_wdir |      1685           1        1684    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g3_wdir |      1685           0        1685    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g4_wdir |      1685           4        1681     |
           countyfip |      1685        1685           0    *|
 countyfip#c.g5_wdir |      1685           1        1684    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g6_wdir |      1685           1        1684    ?|
      fe_countyqtroy |      6676        6676           0    *|
             fe_styr |       631           0         631    ?|
-------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(c.g1_wdir##i.countyfip c.g2_wdir##i.countyf
> ip c.g3_wdir##i.countyfip c.g4_wdir##i.countyfip c.g5_wdir##i.countyfip c.g6_wdir##i.countyfip fe_countyqtroy fe_styr) cluster(c
> ountyfip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 581 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    11.87
                                                      Prob > F      =   0.0006
Total (centered) SS     =  1.50439e+13                Centered R2   =  -0.0162
Total (uncentered) SS   =  1.50439e+13                Uncentered R2 =  -0.0162
Residual SS             =  1.52875e+13                Root MSE      =    15456

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1768.037   513.2517    -3.44   0.001    -2774.716   -761.3591
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             44.149
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2017.807
                         (Kleibergen-Paap rk Wald F statistic):         49.363
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------------+
         Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------------+---------------------------------------|
           countyfip |      1685        1685           0    *|
 countyfip#c.g1_wdir |      1685           2        1683    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g2_wdir |      1685           1        1684    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g3_wdir |      1685           0        1685    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g4_wdir |      1685           4        1681     |
           countyfip |      1685        1685           0    *|
 countyfip#c.g5_wdir |      1685           1        1684    ?|
           countyfip |      1685        1685           0    *|
 countyfip#c.g6_wdir |      1685           1        1684    ?|
      fe_countyqtroy |      6676        6676           0    *|
             fe_styr |       631           0         631    ?|
-------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(c.g1_wdir##i.countyfip c.g2_wdir##i.countyfip c.g3_wd
> ir##i.countyfip c.g4_wdir##i.countyfip c.g5_wdir##i.countyfip c.g6_wdir##i.countyfip fe_countyqtroy fe_styr) cluster(countyfip f
> e_stqtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 581 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =    11.75
                                                      Prob > F      =   0.0006
Total (centered) SS     =  4.79712e+12                Centered R2   =  -0.0098
Total (uncentered) SS   =  4.79712e+12                Uncentered R2 =  -0.0098
Residual SS             =  4.84415e+12                Root MSE      =     8670

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -777.9547   226.9346    -3.43   0.001    -1223.058   -332.8514
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             44.112
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             1993.508
                         (Kleibergen-Paap rk Wald F statistic):         49.886
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------------+
         Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------------+---------------------------------------|
           countyfip |      1686        1686           0    *|
 countyfip#c.g1_wdir |      1686           2        1684    ?|
           countyfip |      1686        1686           0    *|
 countyfip#c.g2_wdir |      1686           1        1685    ?|
           countyfip |      1686        1686           0    *|
 countyfip#c.g3_wdir |      1686           0        1686    ?|
           countyfip |      1686        1686           0    *|
 countyfip#c.g4_wdir |      1686           4        1682     |
           countyfip |      1686        1686           0    *|
 countyfip#c.g5_wdir |      1686           2        1684    ?|
           countyfip |      1686        1686           0    *|
 countyfip#c.g6_wdir |      1686           1        1685    ?|
      fe_countyqtroy |      6677        6677           0    *|
             fe_styr |       637           0         637    ?|
-------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +---------------------------------------------------------+
     |              var      earnings   employment         lfp |
     |---------------------------------------------------------|
  1. | KleibergenPaap_F        50.023                          |
  2. |                                                         |
  3. |        pm25_coef   -123.416***   -1768.0***   -778.0*** |
  4. |      pm25_stderr      (29.247)      (513.3)     (226.9) |
  5. |                                                         |
  6. |     observations        74,725       74,725      75,193 |
     +---------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_b3.dta saved
.                 restore 
.                         
.         }

.         
.         ** panel c, first row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros
> )
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    25.57
                                                      Prob > F      =   0.0000
Total (centered) SS     =   8762876942                Centered R2   =  -0.1232
Total (uncentered) SS   =   8762876942                Uncentered R2 =  -0.1232
Residual SS             =   9842095618                Root MSE      =    364.5

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -103.0771   20.38598    -5.06   0.000    -143.0616   -63.09256
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.437
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2883.998
                         (Kleibergen-Paap rk Wald F statistic):         71.833
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyfip f
> e_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    16.20
                                                      Prob > F      =   0.0001
Total (centered) SS     =  1.77722e+13                Centered R2   =  -0.0154
Total (uncentered) SS   =  1.77722e+13                Uncentered R2 =  -0.0154
Residual SS             =  1.80451e+13                Root MSE      =    15606

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1750.101   434.8435    -4.02   0.000    -2602.992   -897.2105
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.204
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2892.275
                         (Kleibergen-Paap rk Wald F statistic):         71.151
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =    18.85
                                                      Prob > F      =   0.0000
Total (centered) SS     =  6.03036e+12                Centered R2   =  -0.0101
Total (uncentered) SS   =  6.03036e+12                Uncentered R2 =  -0.0101
Residual SS             =  6.09110e+12                Root MSE      =     9039

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -790.8584   182.1385    -4.34   0.000      -1148.1    -433.617
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.381
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2857.845
                         (Kleibergen-Paap rk Wald F statistic):         71.695
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           0         637     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +---------------------------------------------------------+
     |              var      earnings   employment         lfp |
     |---------------------------------------------------------|
  1. | KleibergenPaap_F        71.833                          |
  2. |                                                         |
  3. |        pm25_coef   -103.077***   -1750.1***   -790.9*** |
  4. |      pm25_stderr      (20.386)      (434.8)     (182.1) |
  5. |                                                         |
  6. |     observations        74,725       74,725      75,193 |
     +---------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c1.dta saved
.                 restore 
.                                 
.         }

.         
.         ** panel c, second row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_divisionyr) cluster(countyfip fe_s
> tqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    25.72
                                                      Prob > F      =   0.0000
Total (centered) SS     =   9053870277                Centered R2   =  -0.1604
Total (uncentered) SS   =   9053870277                Uncentered R2 =  -0.1604
Residual SS             =  1.05060e+10                Root MSE      =    375.3

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -110.3243   21.75322    -5.07   0.000    -152.9905   -67.65813
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             46.244
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2497.271
                         (Kleibergen-Paap rk Wald F statistic):         62.535
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
  fe_divisionyr |       117           0         117     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_divisionyr) cluster(count
> yfip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    15.11
                                                      Prob > F      =   0.0001
Total (centered) SS     =  1.91975e+13                Centered R2   =  -0.0318
Total (uncentered) SS   =  1.91975e+13                Uncentered R2 =  -0.0318
Residual SS             =  1.98072e+13                Root MSE      =    16294

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -2183.664   561.7979    -3.89   0.000    -3285.559   -1081.768
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             45.833
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2494.778
                         (Kleibergen-Paap rk Wald F statistic):         61.715
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
  fe_divisionyr |       117           0         117     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_divisionyr) cluster(countyfip fe_st
> qtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 6 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =     3.34
                                                      Prob > F      =   0.0677
Total (centered) SS     =  6.99033e+12                Centered R2   =  -0.0035
Total (uncentered) SS   =  6.99033e+12                Uncentered R2 =  -0.0035
Residual SS             =  7.01448e+12                Root MSE      =     9666

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -637.6759   348.8533    -1.83   0.068    -1321.907    46.55562
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             46.122
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2469.697
                         (Kleibergen-Paap rk Wald F statistic):         62.228
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
  fe_divisionyr |       117           0         117     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +-------------------------------------------------------+
     |              var      earnings   employment       lfp |
     |-------------------------------------------------------|
  1. | KleibergenPaap_F        62.535                        |
  2. |                                                       |
  3. |        pm25_coef   -110.324***   -2183.7***   -637.7* |
  4. |      pm25_stderr      (21.753)      (561.8)   (348.9) |
  5. |                                                       |
  6. |     observations        74,725       74,725    75,193 |
     +-------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c2.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel c, third row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_regionyr) cluster(countyfip fe_stq
> tros)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 6 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    26.27
                                                      Prob > F      =   0.0000
Total (centered) SS     =   9129009833                Centered R2   =  -0.1462
Total (uncentered) SS   =   9129009833                Uncentered R2 =  -0.1462
Residual SS             =  1.04635e+10                Root MSE      =    374.3

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -105.0404    20.4922    -5.13   0.000    -145.2332    -64.8475
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             47.153
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2600.113
                         (Kleibergen-Paap rk Wald F statistic):         61.702
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
    fe_regionyr |        52           0          52     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_regionyr) cluster(countyf
> ip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 6 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    13.52
                                                      Prob > F      =   0.0002
Total (centered) SS     =  1.97391e+13                Centered R2   =  -0.0324
Total (uncentered) SS   =  1.97391e+13                Uncentered R2 =  -0.0324
Residual SS             =  2.03792e+13                Root MSE      =    16520

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -2191.797   596.0809    -3.68   0.000    -3360.934   -1022.659
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             46.759
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2588.234
                         (Kleibergen-Paap rk Wald F statistic):         61.072
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
    fe_regionyr |        52           0          52     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_regionyr) cluster(countyfip fe_stqt
> ros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =     3.18
                                                      Prob > F      =   0.0746
Total (centered) SS     =  7.29919e+12                Centered R2   =  -0.0047
Total (uncentered) SS   =  7.29919e+12                Uncentered R2 =  -0.0047
Residual SS             =  7.33365e+12                Root MSE      =     9879

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -715.3238   400.9494    -1.78   0.075    -1501.735    71.08742
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             47.145
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2582.702
                         (Kleibergen-Paap rk Wald F statistic):         61.607
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
    fe_regionyr |        52           0          52     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +-------------------------------------------------------+
     |              var      earnings   employment       lfp |
     |-------------------------------------------------------|
  1. | KleibergenPaap_F        61.702                        |
  2. |                                                       |
  3. |        pm25_coef   -105.040***   -2191.8***   -715.3* |
  4. |      pm25_stderr      (20.492)      (596.1)   (400.9) |
  5. |                                                       |
  6. |     observations        74,725       74,725    75,193 |
     +-------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c3.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel c, fourth row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy rfrnc_yr) cluster(countyfip fe_stqtro
> s)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    18.26
                                                      Prob > F      =   0.0000
Total (centered) SS     =   9261574242                Centered R2   =  -0.1092
Total (uncentered) SS   =   9261574242                Uncentered R2 =  -0.1092
Residual SS             =  1.02728e+10                Root MSE      =    370.8

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -85.36834   19.97703    -4.27   0.000    -124.5508   -46.18591
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             36.398
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2913.796
                         (Kleibergen-Paap rk Wald F statistic):         47.817
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
       rfrnc_yr |        13           0          13     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy rfrnc_yr) cluster(countyfip 
> fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    13.49
                                                      Prob > F      =   0.0002
Total (centered) SS     =  2.05515e+13                Centered R2   =  -0.0558
Total (uncentered) SS   =  2.05515e+13                Uncentered R2 =  -0.0558
Residual SS             =  2.16986e+13                Root MSE      =    17042

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -2607.633   710.0484    -3.67   0.000    -4000.303   -1214.963
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             35.960
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2902.062
                         (Kleibergen-Paap rk Wald F statistic):         47.151
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
       rfrnc_yr |        13           0          13     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy rfrnc_yr) cluster(countyfip fe_stqtros
> )
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =     0.00
                                                      Prob > F      =   0.9511
Total (centered) SS     =  7.56972e+12                Centered R2   =   0.0000
Total (uncentered) SS   =  7.56972e+12                Uncentered R2 =   0.0000
Residual SS             =  7.56943e+12                Root MSE      =    10034

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -25.71679   418.8625    -0.06   0.951    -847.2623    795.8287
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             36.272
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2886.895
                         (Kleibergen-Paap rk Wald F statistic):         47.660
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
       rfrnc_yr |        13           0          13     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +------------------------------------------------------+
     |              var     earnings   employment       lfp |
     |------------------------------------------------------|
  1. | KleibergenPaap_F       47.817                        |
  2. |                                                      |
  3. |        pm25_coef   -85.368***   -2607.6***     -25.7 |
  4. |      pm25_stderr     (19.977)      (710.0)   (418.9) |
  5. |                                                      |
  6. |     observations       74,725       74,725    75,193 |
     +------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_c4.dta saved
.                 restore 
.                 
.         }

.         
.         use "$Rep_smokelabor/1_build/regdata/county_year.dta", clear 

.         
.         local tbl_settings_log format(%6.3f) parentheses(stderr) asterisk()

.         local tbl_settings_pmil format(%6.1f) parentheses(stderr) asterisk()

.         
.         ** panel d, first row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(countyfip fe_styr) cluster(countyfip)
(dropped 55 singleton observations)
(sum of wgt is     3.4824e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1687               Number of obs =    18916
                                                      F(  1,  1686) =     0.08
                                                      Prob > F      =   0.7734
Total (centered) SS     =  1.51002e+10                Centered R2   =   0.0006
Total (uncentered) SS   =  1.51002e+10                Uncentered R2 =   0.0006
Residual SS             =  1.50915e+10                Root MSE      =    908.1

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   41.56233    144.336     0.29   0.773    -241.5343    324.6589
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.449
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              711.705
                         (Kleibergen-Paap rk Wald F statistic):         50.237
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1687        1687           0    *|
     fe_styr |       613           0         613     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(countyfip fe_styr) cluster(countyfip)
(dropped 55 singleton observations)
(sum of wgt is     2.7605e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1687               Number of obs =    18916
                                                      F(  1,  1686) =     0.01
                                                      Prob > F      =   0.9323
Total (centered) SS     =  2.62886e+12                Centered R2   =   0.0001
Total (uncentered) SS   =  2.62886e+12                Uncentered R2 =   0.0001
Residual SS             =  2.62857e+12                Root MSE      =    11984

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   -99.5618   1172.011    -0.08   0.932     -2398.31    2199.187
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.685
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              728.360
                         (Kleibergen-Paap rk Wald F statistic):         52.476
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1687        1687           0    *|
     fe_styr |       613           0         613     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(countyfip fe_styr) cluster(countyfip)
(dropped 54 singleton observations)
(sum of wgt is     3.5101e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1688               Number of obs =    19004
                                                      F(  1,  1687) =     0.76
                                                      Prob > F      =   0.3820
Total (centered) SS     =  1.18311e+12                Centered R2   =  -0.0010
Total (uncentered) SS   =  1.18311e+12                Uncentered R2 =  -0.0010
Residual SS             =  1.18434e+12                Root MSE      =     8026

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -658.9272   753.5808    -0.87   0.382    -2136.979    819.1244
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.448
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              708.951
                         (Kleibergen-Paap rk Wald F statistic):         50.095
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1688        1688           0    *|
     fe_styr |       619           0         619     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +---------------------------------------------------+
     |              var    earnings   employ~t       lfp |
     |---------------------------------------------------|
  1. | KleibergenPaap_F      50.237                      |
  2. |                                                   |
  3. |        pm25_coef      41.562      -99.6    -658.9 |
  4. |      pm25_stderr   (144.336)   (1172.0)   (753.6) |
  5. |                                                   |
  6. |     observations      18,916     18,916    19,004 |
     +---------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d1.dta saved
.                 restore 
.                                 
.         }

.         
.         ** panel d, second row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(countyfip fe_divisionyr) cluster(countyfip)
(dropped 37 singleton observations)
(sum of wgt is     3.4917e+09)
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1688               Number of obs =    18934
                                                      F(  1,  1687) =     3.72
                                                      Prob > F      =   0.0540
Total (centered) SS     =  1.77920e+10                Centered R2   =  -0.0979
Total (uncentered) SS   =  1.77920e+10                Uncentered R2 =  -0.0979
Residual SS             =  1.95335e+10                Root MSE      =     1019

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -357.4569   185.3477    -1.93   0.054    -720.9925    6.078751
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             14.902
                                                   Chi-sq(1) P-val =    0.0001
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              215.646
                         (Kleibergen-Paap rk Wald F statistic):         20.051
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------+
   Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------+---------------------------------------|
     countyfip |      1688        1688           0    *|
 fe_divisionyr |       117           0         117     |
-------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(countyfip fe_divisionyr) cluster(countyfip)
(dropped 37 singleton observations)
(sum of wgt is     2.7683e+09)
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1688               Number of obs =    18934
                                                      F(  1,  1687) =     4.26
                                                      Prob > F      =   0.0392
Total (centered) SS     =  2.98461e+12                Centered R2   =  -0.0989
Total (uncentered) SS   =  2.98461e+12                Uncentered R2 =  -0.0989
Residual SS             =  3.27981e+12                Root MSE      =    13203

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -4850.989   2350.055    -2.06   0.039    -9460.319   -241.6584
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             14.406
                                                   Chi-sq(1) P-val =    0.0001
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              211.924
                         (Kleibergen-Paap rk Wald F statistic):         19.387
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------+
   Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------+---------------------------------------|
     countyfip |      1688        1688           0    *|
 fe_divisionyr |       117           0         117     |
-------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(countyfip fe_divisionyr) cluster(countyfip)
(dropped 36 singleton observations)
(sum of wgt is     3.5194e+09)
(MWFE estimator converged in 6 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1689               Number of obs =    19022
                                                      F(  1,  1688) =     0.72
                                                      Prob > F      =   0.3955
Total (centered) SS     =  1.42389e+12                Centered R2   =  -0.0167
Total (uncentered) SS   =  1.42389e+12                Uncentered R2 =  -0.0167
Residual SS             =  1.44760e+12                Root MSE      =     8751

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   1048.662   1233.921     0.85   0.396    -1371.514    3468.838
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             14.684
                                                   Chi-sq(1) P-val =    0.0001
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              211.348
                         (Kleibergen-Paap rk Wald F statistic):         19.739
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------+
   Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------+---------------------------------------|
     countyfip |      1689        1689           0    *|
 fe_divisionyr |       117           0         117     |
-------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +-----------------------------------------------------+
     |              var    earnings   employm~t        lfp |
     |-----------------------------------------------------|
  1. | KleibergenPaap_F      20.051                        |
  2. |                                                     |
  3. |        pm25_coef   -357.457*   -4851.0**     1048.7 |
  4. |      pm25_stderr   (185.348)    (2350.1)   (1233.9) |
  5. |                                                     |
  6. |     observations      18,934      18,934     19,022 |
     +-----------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d2.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel d, third row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(countyfip fe_regionyr) cluster(countyfip)
(dropped 37 singleton observations)
(sum of wgt is     3.4917e+09)
(MWFE estimator converged in 6 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1688               Number of obs =    18934
                                                      F(  1,  1687) =     3.64
                                                      Prob > F      =   0.0567
Total (centered) SS     =  1.81984e+10                Centered R2   =  -0.0397
Total (uncentered) SS   =  1.81984e+10                Uncentered R2 =  -0.0397
Residual SS             =  1.89205e+10                Root MSE      =     1001

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -219.8093     115.26    -1.91   0.057    -445.8769    6.258388
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             20.221
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              346.131
                         (Kleibergen-Paap rk Wald F statistic):         25.667
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1688        1688           0    *|
 fe_regionyr |        52           0          52     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(countyfip fe_regionyr) cluster(countyfip)
(dropped 37 singleton observations)
(sum of wgt is     2.7683e+09)
(MWFE estimator converged in 6 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1688               Number of obs =    18934
                                                      F(  1,  1687) =     4.51
                                                      Prob > F      =   0.0338
Total (centered) SS     =  3.12121e+12                Centered R2   =  -0.0700
Total (uncentered) SS   =  3.12121e+12                Uncentered R2 =  -0.0700
Residual SS             =  3.33965e+12                Root MSE      =    13300

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -3929.231   1849.608    -2.12   0.034    -7556.999   -301.4622
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.804
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              333.796
                         (Kleibergen-Paap rk Wald F statistic):         25.064
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1688        1688           0    *|
 fe_regionyr |        52           0          52     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(countyfip fe_regionyr) cluster(countyfip)
(dropped 36 singleton observations)
(sum of wgt is     3.5194e+09)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1689               Number of obs =    19022
                                                      F(  1,  1688) =     0.03
                                                      Prob > F      =   0.8542
Total (centered) SS     =  1.50168e+12                Centered R2   =   0.0011
Total (uncentered) SS   =  1.50168e+12                Uncentered R2 =   0.0011
Residual SS             =  1.50004e+12                Root MSE      =     8893

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -171.9036   935.3844    -0.18   0.854    -2006.539    1662.732
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             20.287
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              346.510
                         (Kleibergen-Paap rk Wald F statistic):         25.753
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1689        1689           0    *|
 fe_regionyr |        52           0          52     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +----------------------------------------------------+
     |              var    earnings   employm~t       lfp |
     |----------------------------------------------------|
  1. | KleibergenPaap_F      25.667                       |
  2. |                                                    |
  3. |        pm25_coef   -219.809*   -3929.2**    -171.9 |
  4. |      pm25_stderr   (115.260)    (1849.6)   (935.4) |
  5. |                                                    |
  6. |     observations      18,934      18,934    19,022 |
     +----------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d3.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel d, fourth row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(countyfip rfrnc_yr) cluster(countyfip)
(dropped 37 singleton observations)
(sum of wgt is     3.4917e+09)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1688               Number of obs =    18934
                                                      F(  1,  1687) =     0.03
                                                      Prob > F      =   0.8684
Total (centered) SS     =  1.87877e+10                Centered R2   =   0.0010
Total (uncentered) SS   =  1.87877e+10                Uncentered R2 =   0.0010
Residual SS             =  1.87694e+10                Root MSE      =      996

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   14.79889   89.28764     0.17   0.868    -160.3273    189.9251
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             30.395
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              646.388
                         (Kleibergen-Paap rk Wald F statistic):         42.072
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1688        1688           0    *|
    rfrnc_yr |        13           0          13     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(countyfip rfrnc_yr) cluster(countyfip)
(dropped 37 singleton observations)
(sum of wgt is     2.7683e+09)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1688               Number of obs =    18934
                                                      F(  1,  1687) =     9.61
                                                      Prob > F      =   0.0020
Total (centered) SS     =  3.32552e+12                Centered R2   =  -0.1281
Total (uncentered) SS   =  3.32552e+12                Uncentered R2 =  -0.1281
Residual SS             =  3.75155e+12                Root MSE      =    14081

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -4585.469   1479.187    -3.10   0.002    -7486.704   -1684.233
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             29.961
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              636.495
                         (Kleibergen-Paap rk Wald F statistic):         41.763
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1688        1688           0    *|
    rfrnc_yr |        13           0          13     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(countyfip rfrnc_yr) cluster(countyfip)
(dropped 36 singleton observations)
(sum of wgt is     3.5194e+09)
(MWFE estimator converged in 5 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1689               Number of obs =    19022
                                                      F(  1,  1688) =     7.41
                                                      Prob > F      =   0.0066
Total (centered) SS     =  1.56954e+12                Centered R2   =  -0.0466
Total (uncentered) SS   =  1.56954e+12                Uncentered R2 =  -0.0466
Residual SS             =  1.64271e+12                Root MSE      =     9296

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   2093.055   768.9976     2.72   0.007     584.7659    3601.344
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             30.031
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              638.222
                         (Kleibergen-Paap rk Wald F statistic):         41.569
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
   countyfip |      1689        1689           0    *|
    rfrnc_yr |        13           0          13     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +------------------------------------------------------+
     |              var   earnings   employment         lfp |
     |------------------------------------------------------|
  1. | KleibergenPaap_F     42.072                          |
  2. |                                                      |
  3. |        pm25_coef     14.799   -4585.5***   2093.1*** |
  4. |      pm25_stderr   (89.288)     (1479.2)     (769.0) |
  5. |                                                      |
  6. |     observations     18,934       18,934      19,022 |
     +------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_d4.dta saved
.                 restore 
.                 
.                 
.         }

.         
.         use "$Rep_smokelabor/1_build/regdata/county_quarter.dta", clear 

.         
.         local tbl_settings_log format(%6.3f) parentheses(stderr) asterisk()

.         local tbl_settings_pmil format(%6.1f) parentheses(stderr) asterisk()

.         
.         ** panel e, first row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr i.fe_countyqtroy##c.rfrnc_yr) c
> luster(countyfip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.3933e+10)
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74769
Number of clusters (fe_stqtros) =   2524              F(  1,  1684) =    11.25
                                                      Prob > F      =   0.0008
Total (centered) SS     =   7374389449                Centered R2   =  -0.0124
Total (uncentered) SS   =   7374389449                Uncentered R2 =  -0.0124
Residual SS             =   7465932361                Root MSE      =    332.7

------------------------------------------------------------------------------
             |               Robust
pc_qwi_pay~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -43.57163   12.98857    -3.35   0.001    -69.04707    -18.0962
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             87.128
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3279.785
                         (Kleibergen-Paap rk Wald F statistic):        115.849
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------------------+
               Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------------------+---------------------------------------|
            fe_countyqtroy |      6676        6676           0    *|
                   fe_styr |       632           0         632     |
            fe_countyqtroy |      6676        6676           0    *|
 fe_countyqtroy#c.rfrnc_yr |      6676           0        6676    ?|
-------------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta saved
.                 
.                 * emp 
.                 ivreghdfe pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr i.fe_countyqtroy##c.rf
> rnc_yr) cluster(countyfip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1046e+10)
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74769
Number of clusters (fe_stqtros) =   2524              F(  1,  1684) =    17.80
                                                      Prob > F      =   0.0000
Total (centered) SS     =  1.56577e+13                Centered R2   =  -0.0043
Total (uncentered) SS   =  1.56577e+13                Uncentered R2 =  -0.0043
Residual SS             =  1.57258e+13                Root MSE      =    15268

------------------------------------------------------------------------------
             |               Robust
pmil_qwi_e~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -847.9991   200.9768    -4.22   0.000     -1242.19   -453.8085
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             87.458
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3306.521
                         (Kleibergen-Paap rk Wald F statistic):        115.860
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------------------+
               Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------------------+---------------------------------------|
            fe_countyqtroy |      6676        6676           0    *|
                   fe_styr |       632           0         632     |
            fe_countyqtroy |      6676        6676           0    *|
 fe_countyqtroy#c.rfrnc_yr |      6676           0        6676    ?|
-------------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta saved
.                 
.                 * lfp
.                 ivreghdfe pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr i.fe_countyqtroy##c.rfrnc_yr) cl
> uster(countyfip fe_stqtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4030e+10)
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75207
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =     9.80
                                                      Prob > F      =   0.0018
Total (centered) SS     =  5.40989e+12                Centered R2   =  -0.0037
Total (uncentered) SS   =  5.40989e+12                Uncentered R2 =  -0.0037
Residual SS             =  5.43003e+12                Root MSE      =     8943

------------------------------------------------------------------------------
             |               Robust
pmil_lau_lfp |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -492.6081   157.3368    -3.13   0.002    -801.2042   -184.0121
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             87.629
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3299.688
                         (Kleibergen-Paap rk Wald F statistic):        116.737
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-------------------------------------------------------------------+
               Absorbed FE | Categories  - Redundant  = Num. Coefs |
---------------------------+---------------------------------------|
            fe_countyqtroy |      6677        6677           0    *|
                   fe_styr |       637           0         637     |
            fe_countyqtroy |      6677        6677           0    *|
 fe_countyqtroy#c.rfrnc_yr |      6677           0        6677    ?|
-------------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +-------------------------------------------------------+
     |              var     earnings   employm~t         lfp |
     |-------------------------------------------------------|
  1. | KleibergenPaap_F      115.849                         |
  2. |                                                       |
  3. |        pm25_coef   -43.572***   -848.0***   -492.6*** |
  4. |      pm25_stderr     (12.989)     (201.0)     (157.3) |
  5. |                                                       |
  6. |     observations       74,769      74,769      75,207 |
     +-------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e1.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel e, second row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.3933e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74769
Number of clusters (fe_stqtros) =   2524              F(  1,  1684) =     1.56
                                                      Prob > F      =   0.2119
Total (centered) SS     =  2.74972e+10                Centered R2   =  -0.0056
Total (uncentered) SS   =  2.74972e+10                Uncentered R2 =  -0.0056
Residual SS             =  2.76517e+10                Root MSE      =    610.7

------------------------------------------------------------------------------
             |               Robust
pc_qwi_pay~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -27.72126   22.19981    -1.25   0.212    -71.26339    15.82088
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.534
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2880.706
                         (Kleibergen-Paap rk Wald F statistic):         71.949
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       632           0         632     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta saved
.                 
.                 * emp 
.                 ivreghdfe pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyfip fe_
> stqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1046e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74769
Number of clusters (fe_stqtros) =   2524              F(  1,  1684) =     9.05
                                                      Prob > F      =   0.0027
Total (centered) SS     =  3.59610e+13                Centered R2   =  -0.0001
Total (uncentered) SS   =  3.59610e+13                Uncentered R2 =  -0.0001
Residual SS             =  3.59630e+13                Root MSE      =    22025

------------------------------------------------------------------------------
             |               Robust
pmil_qwi_e~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   206.9357   68.80363     3.01   0.003      71.9861    341.8854
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.300
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2888.734
                         (Kleibergen-Paap rk Wald F statistic):         71.263
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       632           0         632     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta saved
.                 
.                 * lfp
.                 ivreghdfe pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4030e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1686               Number of obs =    75207
Number of clusters (fe_stqtros) =   2548              F(  1,  1685) =    38.18
                                                      Prob > F      =   0.0000
Total (centered) SS     =  1.29473e+13                Centered R2   =  -0.0029
Total (uncentered) SS   =  1.29473e+13                Uncentered R2 =  -0.0029
Residual SS             =  1.29842e+13                Root MSE      =    13196

------------------------------------------------------------------------------
             |               Robust
pmil_lau_lfp |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -451.9397   73.14329    -6.18   0.000    -595.4009   -308.4784
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.386
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2858.483
                         (Kleibergen-Paap rk Wald F statistic):         71.700
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           0         637     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +----------------------------------------------------+
     |              var   earnings   employ~t         lfp |
     |----------------------------------------------------|
  1. | KleibergenPaap_F     71.949                        |
  2. |                                                    |
  3. |        pm25_coef    -27.721   206.9***   -451.9*** |
  4. |      pm25_stderr   (22.200)     (68.8)      (73.1) |
  5. |                                                    |
  6. |     observations     74,769     74,769      75,207 |
     +----------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e2.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel e, third row 
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (d_pm25=d_hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip fe_stq
> tros)
(dropped 109 singleton observations)
(sum of wgt is     1.2724e+10)
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1653               Number of obs =    67104
Number of clusters (fe_stqtros) =   2328              F(  1,  1652) =    19.87
                                                      Prob > F      =   0.0000
Total (centered) SS     =   7846626169                Centered R2   =  -0.0802
Total (uncentered) SS   =   7846626169                Uncentered R2 =  -0.0802
Residual SS             =   8475811624                Root MSE      =      357

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      d_pm25 |  -65.93967   14.79409    -4.46   0.000     -94.9568   -36.92254
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             68.791
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3228.146
                         (Kleibergen-Paap rk Wald F statistic):         97.851
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         d_pm25
Excluded instruments: d_hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6566        6566           0    *|
        fe_styr |       583           0         583     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (d_pm25=d_hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyf
> ip fe_stqtros)
(dropped 109 singleton observations)
(sum of wgt is     1.0097e+10)
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1653               Number of obs =    67104
Number of clusters (fe_stqtros) =   2328              F(  1,  1652) =    13.47
                                                      Prob > F      =   0.0003
Total (centered) SS     =  1.57213e+13                Centered R2   =  -0.0156
Total (uncentered) SS   =  1.57213e+13                Uncentered R2 =  -0.0156
Residual SS             =  1.59669e+13                Root MSE      =    15493

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      d_pm25 |  -1069.632   291.4834    -3.67   0.000    -1641.348   -497.9164
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             68.931
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3247.509
                         (Kleibergen-Paap rk Wald F statistic):         97.786
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         d_pm25
Excluded instruments: d_hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6566        6566           0    *|
        fe_styr |       583           0         583     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (d_pm25=d_hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqt
> ros)
(dropped 110 singleton observations)
(sum of wgt is     1.2822e+10)
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1653               Number of obs =    67536
Number of clusters (fe_stqtros) =   2352              F(  1,  1652) =     9.11
                                                      Prob > F      =   0.0026
Total (centered) SS     =  5.36262e+12                Centered R2   =  -0.0055
Total (uncentered) SS   =  5.36262e+12                Uncentered R2 =  -0.0055
Residual SS             =  5.39208e+12                Root MSE      =     8975

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      d_pm25 |  -462.7716    153.326    -3.02   0.003    -763.5054   -162.0377
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             69.682
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             3282.550
                         (Kleibergen-Paap rk Wald F statistic):         99.158
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         d_pm25
Excluded instruments: d_hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6566        6566           0    *|
        fe_styr |       588           0         588     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +--------------------------------------------------------+
     |              var     earnings   employment         lfp |
     |--------------------------------------------------------|
  1. | KleibergenPaap_F       97.851                          |
  2. |                                                        |
  3. |      d_pm25_coef   -65.940***   -1069.6***   -462.8*** |
  4. |    d_pm25_stderr     (14.794)      (291.5)     (153.3) |
  5. |                                                        |
  6. |     observations       67,104       67,104      67,536 |
     +--------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_e3.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel f, first row
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip fe_divisio
> nqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_divisionqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_divisionqtros) =    468        F(  1,   467) =    13.94
                                                      Prob > F      =   0.0002
Total (centered) SS     =   8762876942                Centered R2   =  -0.1232
Total (uncentered) SS   =   8762876942                Uncentered R2 =  -0.1232
Residual SS             =   9842095618                Root MSE      =    364.5

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -103.0771   27.61047    -3.73   0.000    -157.3332   -48.82096
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             26.224
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2883.998
                         (Kleibergen-Paap rk Wald F statistic):         35.929
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyfip f
> e_divisionqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_divisionqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_divisionqtros) =    468        F(  1,   467) =     5.28
                                                      Prob > F      =   0.0220
Total (centered) SS     =  1.77722e+13                Centered R2   =  -0.0154
Total (uncentered) SS   =  1.77722e+13                Uncentered R2 =  -0.0154
Residual SS             =  1.80451e+13                Root MSE      =    15606

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1750.101   761.5135    -2.30   0.022    -3246.518   -253.6838
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             26.041
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2892.275
                         (Kleibergen-Paap rk Wald F statistic):         35.545
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_division
> qtros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_divisionqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_divisionqtros) =    468        F(  1,   467) =     9.05
                                                      Prob > F      =   0.0028
Total (centered) SS     =  6.03036e+12                Centered R2   =  -0.0101
Total (uncentered) SS   =  6.03036e+12                Uncentered R2 =  -0.0101
Residual SS             =  6.09110e+12                Root MSE      =     9039

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -790.8584   262.8667    -3.01   0.003    -1307.406   -274.3104
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             26.151
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2857.845
                         (Kleibergen-Paap rk Wald F statistic):         35.783
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           0         637     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +--------------------------------------------------------+
     |              var      earnings   employm~t         lfp |
     |--------------------------------------------------------|
  1. | KleibergenPaap_F        35.929                         |
  2. |                                                        |
  3. |        pm25_coef   -103.077***   -1750.1**   -790.9*** |
  4. |      pm25_stderr      (27.610)     (761.5)     (262.9) |
  5. |                                                        |
  6. |     observations        74,725      74,725      75,193 |
     +--------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f1.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel f, second row
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip fe_regionq
> tros)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_regionqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_regionqtros) =    208          F(  1,   207) =     8.77
                                                      Prob > F      =   0.0034
Total (centered) SS     =   8762876942                Centered R2   =  -0.1232
Total (uncentered) SS   =   8762876942                Uncentered R2 =  -0.1232
Residual SS             =   9842095618                Root MSE      =    364.5

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -103.0771   34.80213    -2.96   0.003    -171.6891   -34.46503
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             17.328
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2883.998
                         (Kleibergen-Paap rk Wald F statistic):         26.282
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyfip f
> e_regionqtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_regionqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_regionqtros) =    208          F(  1,   207) =     2.99
                                                      Prob > F      =   0.0853
Total (centered) SS     =  1.77722e+13                Centered R2   =  -0.0154
Total (uncentered) SS   =  1.77722e+13                Uncentered R2 =  -0.0154
Residual SS             =  1.80451e+13                Root MSE      =    15606

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1750.101   1012.328    -1.73   0.085    -3745.896     245.694
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             17.295
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2892.275
                         (Kleibergen-Paap rk Wald F statistic):         26.095
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_regionqt
> ros)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_regionqtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (fe_regionqtros) =    208          F(  1,   207) =     5.61
                                                      Prob > F      =   0.0188
Total (centered) SS     =  6.03036e+12                Centered R2   =  -0.0101
Total (uncentered) SS   =  6.03036e+12                Uncentered R2 =  -0.0101
Residual SS             =  6.09110e+12                Root MSE      =     9039

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -790.8584   333.9441    -2.37   0.019    -1449.226   -132.4909
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             17.217
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2857.845
                         (Kleibergen-Paap rk Wald F statistic):         25.902
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           0         637     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +------------------------------------------------------+
     |              var      earnings   employ~t        lfp |
     |------------------------------------------------------|
  1. | KleibergenPaap_F        26.282                       |
  2. |                                                      |
  3. |        pm25_coef   -103.077***   -1750.1*   -790.9** |
  4. |      pm25_stderr      (34.802)   (1012.3)    (333.9) |
  5. |                                                      |
  6. |     observations        74,725     74,725     75,193 |
     +------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f2.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel f, third row
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip rfrnc_qtro
> s)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and rfrnc_qtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (rfrnc_qtros) =     52             F(  1,    51) =     5.65
                                                      Prob > F      =   0.0213
Total (centered) SS     =   8762876942                Centered R2   =  -0.1232
Total (uncentered) SS   =   8762876942                Uncentered R2 =  -0.1232
Residual SS             =   9842095618                Root MSE      =    364.5

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -103.0771   43.37154    -2.38   0.021    -190.1491   -16.00508
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             11.238
                                                   Chi-sq(1) P-val =    0.0008
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2883.998
                         (Kleibergen-Paap rk Wald F statistic):         24.959
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyfip r
> frnc_qtros)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and rfrnc_qtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (rfrnc_qtros) =     52             F(  1,    51) =     1.23
                                                      Prob > F      =   0.2731
Total (centered) SS     =  1.77722e+13                Centered R2   =  -0.0154
Total (uncentered) SS   =  1.77722e+13                Uncentered R2 =  -0.0154
Residual SS             =  1.80451e+13                Root MSE      =    15606

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1750.101   1579.744    -1.11   0.273    -4921.569    1421.366
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             11.192
                                                   Chi-sq(1) P-val =    0.0008
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2892.275
                         (Kleibergen-Paap rk Wald F statistic):         24.935
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip rfrnc_qtros
> )
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and rfrnc_qtros

Number of clusters (countyfip) =   1686               Number of obs =    75193
Number of clusters (rfrnc_qtros) =     52             F(  1,    51) =     2.39
                                                      Prob > F      =   0.1280
Total (centered) SS     =  6.03036e+12                Centered R2   =  -0.0101
Total (uncentered) SS   =  6.03036e+12                Uncentered R2 =  -0.0101
Residual SS             =  6.09110e+12                Root MSE      =     9039

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -790.8584   511.1455    -1.55   0.128    -1817.026     235.309
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             11.224
                                                   Chi-sq(1) P-val =    0.0008
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2857.845
                         (Kleibergen-Paap rk Wald F statistic):         24.615
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           0         637     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +----------------------------------------------------+
     |              var     earnings   employ~t       lfp |
     |----------------------------------------------------|
  1. | KleibergenPaap_F       24.959                      |
  2. |                                                    |
  3. |        pm25_coef   -103.077**    -1750.1    -790.9 |
  4. |      pm25_stderr     (43.372)   (1579.7)   (511.1) |
  5. |                                                    |
  6. |     observations       74,725     74,725    75,193 |
     +----------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f3.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel f, fourth row
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(countyfip)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1685               Number of obs =    74725
                                                      F(  1,  1684) =    57.35
                                                      Prob > F      =   0.0000
Total (centered) SS     =   8762876942                Centered R2   =  -0.1232
Total (uncentered) SS   =   8762876942                Uncentered R2 =  -0.1232
Residual SS             =   9842095618                Root MSE      =    364.5

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -103.0771    13.6108    -7.57   0.000    -129.7729   -76.38123
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):            152.237
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2883.998
                         (Kleibergen-Paap rk Wald F statistic):        190.525
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(countyfip)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1685               Number of obs =    74725
                                                      F(  1,  1684) =    56.23
                                                      Prob > F      =   0.0000
Total (centered) SS     =  1.77722e+13                Centered R2   =  -0.0154
Total (uncentered) SS   =  1.77722e+13                Uncentered R2 =  -0.0154
Residual SS             =  1.80451e+13                Root MSE      =    15606

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1750.101   233.3966    -7.50   0.000    -2207.879   -1292.323
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):            150.678
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2892.275
                         (Kleibergen-Paap rk Wald F statistic):        188.795
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip

Number of clusters (countyfip) =   1686               Number of obs =    75193
                                                      F(  1,  1685) =    40.76
                                                      Prob > F      =   0.0000
Total (centered) SS     =  6.03036e+12                Centered R2   =  -0.0101
Total (uncentered) SS   =  6.03036e+12                Uncentered R2 =  -0.0101
Residual SS             =  6.09110e+12                Root MSE      =     9039

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -790.8584    123.867    -6.38   0.000    -1033.808    -547.909
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):            152.283
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2857.845
                         (Kleibergen-Paap rk Wald F statistic):        190.821
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637           0         637     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +---------------------------------------------------------+
     |              var      earnings   employment         lfp |
     |---------------------------------------------------------|
  1. | KleibergenPaap_F       190.525                          |
  2. |                                                         |
  3. |        pm25_coef   -103.077***   -1750.1***   -790.9*** |
  4. |      pm25_stderr      (13.611)      (233.4)     (123.9) |
  5. |                                                         |
  6. |     observations        74,725       74,725      75,193 |
     +---------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f4.dta saved
.                 restore 
.                 
.         }

.         
.         ** panel f, fifth row
.         if 1 {
.                 
.                 * income 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop], a(fe_countyqtroy fe_styr) cluster(statefip)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on statefip

Number of clusters (statefip) =     49                Number of obs =    74725
                                                      F(  1,    48) =     6.69
                                                      Prob > F      =   0.0128
Total (centered) SS     =   8762876942                Centered R2   =  -0.1232
Total (uncentered) SS   =   8762876942                Uncentered R2 =  -0.1232
Residual SS             =   9842095618                Root MSE      =    362.9

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -103.0771   39.84508    -2.59   0.013     -183.191   -22.96318
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             10.558
                                                   Chi-sq(1) P-val =    0.0012
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2908.520
                         (Kleibergen-Paap rk Wald F statistic):         18.175
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631         631           0    *|
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta", addlabel(KleibergenPaap_F, `e(r
> kf)') table(earnings, `tbl_settings_log') replace
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta saved
.                 
.                 * emp 
.                 ivreghdfe d_pmil_qwi_emptotal (pm25=hms_deep) [aw=seer_pop16plus], a(fe_countyqtroy fe_styr) cluster(statefip)
(dropped 187 singleton observations)
(sum of wgt is     1.1026e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on statefip

Number of clusters (statefip) =     49                Number of obs =    74725
                                                      F(  1,    48) =    12.10
                                                      Prob > F      =   0.0011
Total (centered) SS     =  1.77722e+13                Centered R2   =  -0.0154
Total (uncentered) SS   =  1.77722e+13                Uncentered R2 =  -0.0154
Residual SS             =  1.80451e+13                Root MSE      =    15540

------------------------------------------------------------------------------
             |               Robust
d_pmil_qwi~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1750.101   503.0407    -3.48   0.001    -2761.532   -738.6699
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             10.462
                                                   Chi-sq(1) P-val =    0.0012
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2916.868
                         (Kleibergen-Paap rk Wald F statistic):         17.943
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631         631           0    *|
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta", table(employment, `tbl_settings
> _pmil') append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta saved
.                 
.                 * lfp
.                 ivreghdfe d_pmil_lau_lfp (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(statefip)
(dropped 189 singleton observations)
(sum of wgt is     1.4027e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on statefip

Number of clusters (statefip) =     49                Number of obs =    75193
                                                      F(  1,    48) =     7.69
                                                      Prob > F      =   0.0079
Total (centered) SS     =  6.03036e+12                Centered R2   =  -0.0101
Total (uncentered) SS   =  6.03036e+12                Uncentered R2 =  -0.0101
Residual SS             =  6.09110e+12                Root MSE      =     9000

------------------------------------------------------------------------------
             |               Robust
d_pmil_lau~p |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -790.8584   285.1058    -2.77   0.008    -1364.102   -217.6148
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             10.562
                                                   Chi-sq(1) P-val =    0.0012
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2882.224
                         (Kleibergen-Paap rk Wald F statistic):         18.137
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6677        6677           0    *|
        fe_styr |       637         637           0    *|
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.                 regsave using "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta", table(lfp, `tbl_settings_pmil')
>  append
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta saved
.                 
.                 preserve 
.                         use "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta", clear
.                         drop if var == "r2"
(1 observation deleted)
.                         drop if var == "_id"
(0 observations deleted)
.                         drop if strpos(var, "_cons_") > 0
(0 observations deleted)
.                         replace var="observations" if var=="N"
(1 real change made)
.                         ingap 2
.                         ingap 5
.                         list , sep(0)

     +--------------------------------------------------------+
     |              var     earnings   employment         lfp |
     |--------------------------------------------------------|
  1. | KleibergenPaap_F       18.175                          |
  2. |                                                        |
  3. |        pm25_coef   -103.077**   -1750.1***   -790.9*** |
  4. |      pm25_stderr     (39.845)      (503.0)     (285.1) |
  5. |                                                        |
  6. |     observations       74,725       74,725      75,193 |
     +--------------------------------------------------------+
.                         saveold "$Rep_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta", replace
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_tables/appendix_table2_f5.dta saved
.                 restore 
.                 
.         }

.         
.         
.         
. log close
      name:  <unnamed>
       log:  C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_logs/appendix_table2.log
  log type:  text
 closed on:  24 Jun 2022, 07:44:55
----------------------------------------------------------------------------------------------------------------------------------
